Simulation of rill erosion in black soil and albic soil during the snowmelt period

ABSTRACT

Snowmelt-induced rill erosion could bring serious harm for soil quality and agricultural productive conditions of slope farmland in the black soil zone of Northeast China. In this study, we conducted laboratory experiments to investigate the effects of the freeze-thaw (FT) temperature, number of FT cycles, water content, flow rate, and thaw depth on rill morphology and erosion amount in two common soil (black soil and albic soil). The thaw depth obtained the maximum range, which was the primary factor for the width-to-depth ratio of rills in the black soil; whereas, the flow rate obtained the maximum range as the primary factor for rill erosion in black soil and albic soil. The number of FT cycles had a minor effect on rill erosion in the two soils. Under the same conditions, the rill morphology showed a large difference between the two soils, and higher rill erosion occurred in albic soil than black soil. Rill erosion was relatively high in black soil and albic soil when the FT temperature fluctuated around 0°C during freezing-thawing. The water content exhibited a greater effect on rill erosion in black soil than in albic soil. The unthawed frozen layer could promote rill erosion during snowmelt period to some extent. The results could provide some reference for future study snowmelt-induced rill erosion mechanism and preventive measures.

KEYWORDS:

• Rill erosion
• rill morphology
• snowmelt erosion
• freezing-thawing
• snowmelt runoff

Introduction

Soil erosion shows seasonal changes in the majority of middle and high latitudes on Earth and runoff erosion due to spring snowmelt accounts for the vast majority of the annual soil and water loss in part of the freeze-thaw (FT) eroded areas (Demildov et al. 1995). The black soil region of Northeast China is an important commodity grain base in China, where soil freezing and thawing have serious effect during the winter and early spring.

The soil erosion that occurs in the spring thaw period differs from general rainfall erosion. During the thaw period, the surface soil thaws whereas the deep soil freezes, forming an impermeable layer. Consequently, snowmelt water cannot infiltrate into the soil in a timely manner and thus increases the scouring ability of snowmelt runoff (Flanagan & Nearing 1995). In the thaw period, the soil undergoes multiple FT cycles, and the surface soil contains high water content due to immersion in snowmelt water, thus significantly reducing the anti-erosion ability of soil (Yoo et al. 1982). Oztas and Fayetorbay (2003) showed that after FT cycles, soil aggregate stability is reduced by 28.3–51.7% in different soil types with varying soil water content. Fan et al. (2011b) reported a laboratory FT test showing that freezing-thawing strongly affects the cohesion of black soil, with little effect on the internal friction angle. Based on an analysis of five major soil types from rainfed cropland in Northeast China, Zhou et al. (2011) found that the bulk density is decreased, whereas the porosity is increased, after freezing-thawing action, making the soil more prone to erosion.

Rill erosion, which is one of the major forms of slope soil erosion, could change slope morphology significantly. Shi et al. (2009) found that the emergence of rills can exacerbate soil erosion during the generation of snowmelt runoff. Beullens et al. (2014) found that the flow rate affects rill erosion in Belgium and northern France. In a laboratory experiment simulating rill erosion with FT, Gatto (2000) observed the following changes in vegetation-free rectangular rills after two FT cycles: the soil water content increased, whereas the surface soil cohesion decreased; the rills were downcut, and the rill walls collapsed, resulting in a triangular cross-section and significantly affecting rill morphology. Based on a study of a winter rainfall event in Norway, Oygarden (2003) proposed that an incompletely thawed layer is the main cause of a wide range of sheet erosion and rill erosion. Bayard et al. (2005) found that the melting of ice sheets decreases the infiltration capacity and increases the rill erosion of soil during the spring thaw period.

In summary, various factors, including the process of soil FT cycles (FT temperature and number of cycles), the soil water content, the thaw depth, and the flow rate, strongly affect the rill erosion of soil during the spring thaw period. In this study, black soil and albic soil are used to simulate the process of snowmelt erosion in the spring thaw period under laboratory conditions. The effects of five factors, that is, the FT temperature, the number of cycles, the water content, the flow rate, and the thaw depth, on rill erosion and rill morphology in the snowmelt period were studied which provide evidence for the forecast and prevention of rill erosion by snowmelt in the black soil zone of Northeast China during the spring snowmelt period.

Materials and methods

Experimental soil samples

The experiment was conducted using samples of surface soil (0–20 cm) from farmland including two typical soils, that is, black soil and albic soil, under rainfed cultivation in Northeast China. The black soil was taken from the Changshuihe Farm in Erjing Town, Bei’an City, Heilongjiang Province, China. Black soil is mainly distributed in the Jilin and Heilongjiang Provinces, which belong to the North Temperate Zone, with a semi-humid continental monsoon climate. The black soil zone has a cold winter and dry, windy spring and autumn. The average annual temperature varies from 1°C to 6°C, and the average January temperature ranges between −10°C and −20°C, with a significant north-to-south difference. The black soil zone has large diurnal and annual ranges of temperature, and alternate freezing-thawing is significant and long lasting. In the cold winter, soil freezes to a deep depth of approximately 1.3 –3 m, with a surface-freezing period of 150 –180 days.

The albic soil was taken at the junction of Meihekou City, Jilin Province, and Caoshi Town, Qingyuan County, Liaoning Province. Albic soil is mainly distributed in the humid northeast region of Northeast China, and this soil type is primarily found in the eastern part of Heilongjiang and Jilin Provinces. The albic soil zone has a climate characterized by a cold, dry winter and a warm, humid summer. The average annual temperature varies from 4°C to 6°C, and the average temperature in the coldest month ranges between −10°C and −15°C. In the cold winter, soil freezes to a deep depth of approximately 1.5 –2 m, with a surface-freezing period of 150 –170 days.

Based on multiple comparison tests, the bulk densities and saturated water contents of the experimental soils were obtained: 0.9 g/cm³ and 55.76% for the black soil, respectively, and 1.25 g/cm³ and 32.13% for the albic soil, respectively. The mechanical composition of the two soils is shown in Table 1.

Table 1. Mechanical composition of the experimental soils.

Soil texture	Medium sand	Fine sand	Coarse silt	Physical sand	Medium silt	Fine silt	Clay	Physical clay
	1.00–0.25	0.25–0.05	0.05–0.01	>0.01	0.01–0.005	0.005–0.001	<0.001	<0.01
Black soil	0.5	4.2	26.7	31.4	10.3	15.5	42.8	68.6
Albic soil	12.43	2.37	31.92	46.72	7.22	11.51	34.55	53.28

Experimental procedure

The experiment was conducted under laboratory conditions to simulate rill erosion by snowmelt runoff during the spring thaw period. The apparatus used for the simulation of snowmelt runoff erosion comprised the following five connecting components: a water supply tank, a flow rate control valve, a steady flow channel, a runoff slope, and a soil sample. The snowmelt runoff simulation apparatus was horizontally set up at the highest point of the water channel surface on the top of the runoff plot. In order to meet the design requirements, the experimental water was taken from the water supply tank, by controlling the mixing ratio of ice and water, and reading the thermometer, kept the water temperature at 0°C; meanwhile, the temperature and the internal energy of the test water can be kept consistent with that of the snowmelt runoff. The steady flow channel was connected with water supply tank, and the other end was connected with the runoff slope. The runoff slope was on the same plane as the soil surface, which can make the runoff flow evenly and close to the natural state. The runoff slope was 50 cm in length and 20 cm in width. The soil sample was 50 cm in length, 20 cm in width, and 10 cm in height.

As revealed by a survey of erosion data in the thaw period and topographical data at the experimental site, the black soil zone of Northeast China mainly has wave-like rolling hills with small undulations and gentle slopes, generally 1–5°; however, the topography becomes complex with large undulations, and most of the ground slopes 3–8° in the piedmont alluvial and diluvial terrace areas. The slope is even larger in particular regions, creating favorable conditions for soil erosion. This favorable condition is the main cause leading to soil erosion during the spring thaw period. Therefore, the slope was consistently set to 8° for placement of the experimental apparatus. The FT test equipment was manufactured by the Shenyang First Refrigerating Machinery Company (Shenyang, China). The temperature of the FT machine was controlled in the −40°C to 40°C range by a temperature controller.

The soil samples collected in the field were passed through a 5 mm × 5 mm sieve to remove roots and impurities. After the water content measurement, if the water content was lower than the design requirement, then the soil was wetted using a watering can to reach the water content specified in the design and covered with a plastic sheet for 48 h to obtain uniform soil water content. If the measured water content was higher than the design requirement, then the soil sample was spread in a dark, ventilated, dry place, allowing for natural evaporation of water to reach the water content specified in the design before filling the soil bin. When filling the soil, the load required for each layer was calculated according to the bulk density of undisturbed soil, 0.9 and 1.25 g/cm³. The soil was filled in 10 layers, 1 cm per layer, to ensure uniform distribution. After each layer of soil was filled, the soil surface was roughened to ensure the tight connection of different layers. After filling the experimental soil, the sample was placed in the FT machine, and the FT test was conducted for a specific number of FT cycles. When the soil sample completed the required number of FT cycles, a measuring probe of white steel material (0.5 cm diameter) was inserted from the soil surface into the soil mass at 5 min intervals after the last time of thaw until the unthawed frozen layer was reached. The length of the probe in the soil mass was taken as the thaw depth. When the thaw depth met the experimental requirement, the snowmelt scour simulation was started.

Before each round of scour, the flow rate was calibrated according to the orthogonal experimental design. Start experiment when the steady flow channel was filled and overflowed. The experiment simulated upslope snowmelt water evenly scoured toward the downhill. The scoured sediment and runoff were all collected in the collection tank in the lower part of the runoff plot. To reduce the melting of the frozen soil layer by the runoff, the time of the snowmelt runoff simulation was set to 20 min. The process of runoff scour was carefully observed during the experiment. Additionally, the position coordinates (every 5 cm) and depth (every 5 cm) of the rills were recorded after the experiment. Images were acquired using a camera in the experimental process, supplemented with manual recording.

Experimental design

Five influencing factors were selected in the experiment, that is, the FT temperature, the number of cycles, the soil water content, the flow rate, and the soil thaw depth. Each factor was set to five levels. An orthogonal design method was used, without considering the interaction between various factors. The experiment was designed using the L₂₅ (5⁶) orthogonal table. A total of 25 tests were performed in each black soil and albic soil group, with three replications each. The influencing factors and their levels used in the experiment are shown in Table 2.

Table 2. The factors and levels of the orthogonal experiment.

Levels	FT temperature (°C)	Number of cycles (times)	Soil water content (%)	Flow rate (L/min)	Soil thaw depth (cm)
1	−30∼3	5	20	1	1
2	−25∼5	10	25	2	2
3	−10∼7	15	30	3	3
4	−5∼13	20	35	4	4
5	−3∼15	30	40	5	5

Data analysis

The data were analyzed by range analysis using SPSS 19.0. Range analysis is the use of mathematical statistics method to calculate the orthogonal table for each column range and determine the primary and secondary relationship between factors. The level and trend of the effects of the five factors (FT temperature, number of cycles, soil water content, flow rate, and thaw depth) on the rill morphology and rill erosion were compared.

The width-to-depth ratio (WDR) of the rills was chosen to characterize the rill morphology. The WDR can indicate the wide, shallow and narrow, deep degrees of the cross-section of rills. By referring to the research results of rivers, the ζ parameter was chosen to characterize the sectional characteristics of the rills. The calculation formula is as follows:where B (m) is the rill width, and h (m) is the rill depth.

Results and discussion

Rill morphology during snowmelt

A comparison of the rill morphology between the two soils demonstrated that the rill morphology showed a substantial difference between the black soil and the albic soil under the same conditions. Because the black soil and the albic soil were characterized by different structures and properties, their WDRs of rills showed a large difference (Figures 1 and 2). In the black soil, the maximum WDR of rills was obtained in test 13, that is, 8.82; the minimum value was obtained in test 9, that is, 2.18; and the WDR of rills varied between 2.26 and 6.67 in the remaining tests. In the albic soil, the maximum WDR of rills was obtained in test 2, that is, 9.38; the minimum value was obtained in test 4, that is, 3.33; and the WDR of rills varied between 3.53 and 9.35 in the remaining tests. The largest difference in the WDRs of rills between the black soil and the albic soil was obtained in test 15, and the absolute value of the difference was 4.83. In contrast, the same WDR of rills was obtained in test 4 for the black soil and the albic soil, that is, the difference was zero.

Figure 1. Comparison of the WDRs of rills between the black soil and the albic soil.

Figure 2. Comparison of rill erosion between the black soil and the albic soil.

The rill morphology under snowmelt runoff appeared to be different from that generated by rainfall. The formation process of rills by snowmelt was also a thawing process of the frozen soil layer. The thaw depth limited the downward development of rills, whereas runoff promoted the melting of the frozen soil layer. A range analysis was conducted on the WDR of the rills in the black soil, and the results of the factor level mean K showed that the effects of the various factors on the WDR rank as follows: thaw depth > FT temperature > water content > number of cycles > flow rate. In the black soil, there was a decrease followed by an increase in the WDR of rills with increasing flow rate and thaw depth. In contrast, an increase followed by a decrease was found in the WDR with increasing number of cycles. The WDR of rills fluctuated in a decrease–increase–decrease pattern with varying FT temperature and an increase–decrease–increase pattern with increasing water content.

The soil thaw depth showed the greatest effect on the WDR of rills in the black soil. The maximum WDR of rills (6.18) was found at the thaw depth of 1 cm, and the WDR gradually decreased and then increased with increasing thaw depth in the black soil. In terms of the FT temperature, the maximum WDR of rills appeared at −10°C to 7°C, and the corresponding rill erosion also reached the maximum, with a high rate of rill wall collapse. Additionally, the WDR of rills was relatively high in the black soil with the water content of 25%. Moreover, the number of FT cycles affected the WDR of rills to a certain degree. High WDRs of rills were obtained after 5, 10, and 15 FT cycles (4.7, 4.95, and 5.2, respectively). The flow rate showed a minor effect on the WDR of rills in the black soil.

A range analysis was conducted on the WDR of rills in the albic soil, and the results of the factor level mean K showed that the effects of the various factors on the WDR of rills rank as follows: flow rate > water content > FT temperature > thaw depth > number of cycles. In the albic soil, the WDR of rills initially decreased and then increased with increasing flow rate. The WDR fluctuated in an increase–decrease–increase pattern with increasing number of cycles, water content, and thaw depth, whereas it showed a decrease–increase–decrease trend with varying FT temperature.

The flow rate was the most important factor affecting the WDR of rills in the albic soil, and the maximum WDR of rills (5.68) appeared at the flow rate of 1 L/min. With regard to the water content and FT temperature, the maximum WDRs of rills (5.83 and 5.72, respectively) appeared at 25% and −10°C to 7°C, respectively. In terms of the thaw depth, the maximum WDR of rills was obtained at 3 cm. The number of cycles showed a minor effect on the WDR of rills in the albic soil, and the maximum and minimum of the WDR (5.57 and 4.52, respectively) were obtained after 10 and 5 cycles, respectively.

Differences of influencing factors in rill morphology

The effects of five factors on the two soils were also different (Figures 1 and 2). The thaw depth was the primary factor affecting the WDR of rills in the black soil. The black soil obtained a relatively high WDR of rills at the thaw depth of 1 cm. The reason is that the presence of the frozen soil layer hampered the downcut of rills, thus forming a broad shallow rill morphology. However, thaw depth showed less effect on the WDR of rills in the albic soil. Given the poor permeability of albic soil, the frozen soil layer had a minor effect on the rill morphology (Liu et al., 2012). The results verifies the results of Oygarden (2003) that incompletely thawed layer one part acts to alter the cross-sectional shape of the rills, whereas the other part determines the flow rate and rill depth..

The FT temperature and water content showed similar effects on the WDR of rills in the black soil and the albic soil. The effect of the FT temperature for the WDR of rills was greater in the black soil than in the albic soil. In the FT process, the black soil had a lower bulk density and a higher porosity than the albic soil (Zhou et al. 2011). The FT temperature thus affected the black soil more than the albic soil. In the black soil, the WDR of rills exhibited a similar trend as rill erosion, and both parameters reached the maximum values at −10°C to 7°C. In the albic soil, the WDR of rills was relatively high at the FT temperature of −10°C to 7°C and −5°C to 13°C, which is generally consistent with the trend of maximum rill erosion. These results indicate that with high rill erosion, rill bank expansion was more significant than rill downcutting, resulting in a broad shallow rill morphology. Both of the soils obtained the maximum WDR of rills at the initial water content of 20%. With increasing water content, water-stable aggregates gradually increased in the soil (Li & Fan 2014), and the anti-erosion ability of the surface soil was enhanced, leading to the downcutting of the melting frozen layer by rills.

The number of FT cycles showed a minor effect on the WDR of rills in the black soil and the albic soil. The number of FT cycles affected rill erosion by altering the soil structure and physical-mechanical properties through alternating freezing and thawing. With increasing number of FT cycles, the soil bulk density was reduced, whereas the soil porosity was increased, making the soil more prone to erosion (Liu et al. 2009). Because the number of FT cycles had no strong effect on the frozen soil layer, it showed a minor effect on the WDR of rills.

The flow rate showed markedly different effects on the WDR of rills in the black soil and the albic soil. The effect of the flow rate on the WDR of rills was limited in the black soil. Because the soil was more susceptible to erosion after FT cycles, the flow rate had little effect on the rill morphology. However, with increasing flow rate, the WDR of rills was decreased and finally a slight increase, whereas rill erosion was increased. This phenomenon indicates that for the change in rill erosion, rill downcutting contributed more to rill erosion in the albic soil than gully bank expansion. The reason is that the albic soil had a heavy clay texture, which means that soil porosity was increased after FT. With a high initial infiltration rate, runoff accelerated the melting of the underlying frozen layer. For the albic soil, rapid downcutting occurred in the rills, resulting in a narrow deep morphology.

Rill erosion during snowmelt

The rill erosion amount was calculated using the volume method. In brief, the rill length, width, and depth were measured to calculate the rill volume, which was then multiplied by the soil bulk density to obtain the rill erosion. The two soils were compared for rill erosion by snowmelt (Figure 2). The results showed that the rill erosion of the black soil obtained the maximum of 1372.39 g in test 16 and the minimum of 92.32 g in test 24; the rill erosion varied in the 172.80–1197.96 g range for the other tests. The rill erosion of the albic soil obtained the maximum of 3296.94 g in test 16 and the minimum of 323.20 g in test 15; the rill erosion varied in the 382.40–2410.69 g range for the remaining tests. The largest difference in rill erosion between the two soils was found in test 17, where the albic soil showed a 4.98-fold increase in rill erosion compared to the black soil. The smallest difference was found in test 15, where the albic soil showed a 0.36-fold increase in rill erosion compared to the black soil. The rill erosion of the albic soil was significantly higher than that of the black soil under the same conditions. This finding can be attributed to the higher dispersion of the albic soil compared to the black soil (Zhang & Zhang 1988), which can facilitate soil erosion.

A range analysis was conducted on rill erosion in the black soil, and the results of the factor level mean K showed that the effect of the various factors on rill erosion ranked as follows: flow rate > water content > FT temperature > thaw depth > number of cycles. In the black soil, rill erosion linearly increased with increasing flow rate, whereas it gradually decreased with increasing FT cycles. As the thaw depth increased, rill erosion showed an increase followed by a decrease. Moreover,soil erosion fluctuated in a decrease-increase-decrease-increase pattern with increasing FT temperature. With increasing water content, rill erosion showed large fluctuations in an increase–decrease–increase–decrease pattern.

In the black soil, the rill erosion was 202.9 g at the flow rate of 1 L/min; the rill erosion reached 956.68 g (>3-fold increase) at the flow rate of 5 L/min. With regard to the water content, the maximum (824.62 g) and minimum (520.74 g) of rill erosion were obtained at 25% and 40%, respectively. With varying FT temperature, the maximum of rill erosion (780.23 g) was obtained at −10°C to 7°C. The thaw depth and the number of cycles showed minor effects on rill erosion in the black soil. In terms of the thaw depth, the maximum (735.95 g) and minimum (532.89 g) of rill erosion were obtained at 3 and 5 cm, respectively. With regard to the number of cycles, the maximum and minimum soil erosion (505.88 and 706.61 g, respectively) were obtained after 5 and 10 FT cycles, respectively.

A range analysis was conducted on rill erosion in the albic soil, and the results of the factor level mean K showed that the effect of the various factors on rill erosion ranked as follows: flow rate > FT temperature > thaw depth > water content > number of cycles. In the albic soil, soil erosion linearly increased with increasing flow rate. A decrease followed by an increase occurred in soil erosion with increasing number of FT cycles. The opposite trend was found with increasing thaw depth. Moreover, soil erosion fluctuated in a decrease–increase–decrease pattern with increasing FT temperature, whereas it showed an increase–decrease–increase pattern with increasing water content.

In the albic soil, the rill erosion was 564.16 g at the flow rate of 1 L/min, and the erosion increased to 2174.83 g when the flow rate reached 5 L/min. With regard to the other factors, the maximum rill erosion was obtained at the FT temperature of −5°C to 13°C (1739.50 g), the thaw depth of 3 cm (1612.30 g), the water content of 25% (1622.89 g), and the number of cycles of 5 (1552.33 g).

Differences of influencing factors in rill erosion

Comparing the extremes of maximum and minimum values of the black soil and the albic soil, it can be seen that the flow rate has a significant effect on the amount of rill erosion in the two soils (Figure 2). The snowmelt flow rate obtained the maximum range (R), indicating that the flow rate was the primary influencing factor for rill erosion in both of the soils. The rill erosion was increased in both soils with increasing flow rate. The average rates par 1 L/min of increase were 56.16% in the black soil and 44.20% in the albic soil, indicating that the rill erosion of black soil was more sensitive to changes in the flow rate. The flow rate remained relatively high in the soil scoured by snowmelt water after FT. The higher the flow rate was, the larger the runoff, the stronger the sediment carrying capacity, and the more severe the erosion. Moreover, the soil characteristics were altered in the surface soil after FT cycles. The surface soil became loose and produced large amounts of erodible materials, and the large flow rate accelerated the melting of the frozen soil layer. Therefore, a higher flow rate had a more significant effect on slope erosion. The results verify Fan et al. (2011a) that the main factor affecting soil erosion during snowmelt period is the flow rate. This finding is similar to the trend in unthawed soil, in which rill erosion is increased with increasing flow rate (Wang et al. 2014).

The effect of the water content was greater for the black soil erosion than for the albic soil erosion. The saturated water content of black soil (55.76%) is higher than that of albic soil (32.13%). When albic soil reaches water content saturation, the soil pores were completely filled with liquid water, changing degree of the soil structure is similar between soils, and the changes in rill erosion are minor. Therefore, the water content had a greater effect on rill erosion in the black soil. Both of the soils obtained the maximum rill erosion at 25% water content. This result indicates that with the initial water content of 25%, the soil after thawing was characterized by less ice, more pores, and a strong infiltration capacity; this soil was less affected by the incompletely thawed layer, and it obtained a high thawing rate, contributing to the maximum total erosion. Conversely, the minimum rill erosion was obtained in the black soil with the initial water content of 40% and in the albic soil with the initial water content of 35%. Compared to the black soil, the albic soil had a heavier clay texture and a higher initial infiltration rate; however, the infiltration rate was also slowly decreased with increasing water content (Zhou et al. 2011). In the early stage of snowmelt runoff erosion, an impermeable layer was formed by freezing-thawing in soil with a higher water content, which increased the rill erosion. However, with the continuous development of the erosion, the frozen layer restricted the downward channel development, thus affecting rill erosion. In the soil with a low initial water content, the frozen layer had good permeability, allowing a rapid downward development of erosion gullies and thus resulting in high rill erosion.

With regard to the FT temperature, the maximum rill erosion was obtained in the black soil at −10°C to 7°C and in the albic soil at −5°C to 13°C. The rill erosion was relatively low in both soils at −30°C to 3°C and −25°C to 5°C, indicating a lack of a clear relationship between the minimum soil freezing temperature and rill erosion. The rill erosion was relatively high when the temperature fluctuated around 0°C during freezing-thawing.

For both the black soil and the albic soil, the maximum rill erosion was obtained at the thaw depth of 3 cm. The reason is that at a small thaw depth, the presence of the frozen soil layer prevented any further development of rill erosion. At a greater soil thaw depth, the frozen soil layer was farther from the soil surface; water infiltration reduced the surface flow rate and thereby reduced rill erosion.

The effect of the number of FT cycles on soil erosion was achieved by altering the soil structure and physical-mechanical properties through alternate freezing and thawing. The rill erosion was relatively high in both the black soil and the albic soil after 5–10 FT cycles. As the cycle number continued to increase, the rill erosion showed a decreasing trend. With increasing number of FT cycles, the soil bulk density was decreased, the porosity was increased, and the water-stable aggregate content was decreased (Liu et al. 2009). However, the changes in the water-stable aggregate content gradually reached a steady state after seven FT cycles (Li & Fan 2014). Therefore, the rill erosion was tending to stability in the two soils.

Our study showed some preliminary and relatively comprehensive results about the major influencing factors of snowmelt-induced rill erosion in the black soil zone of Northeast China. For both black soil an albic soil, snowmelt runoff is the direct and most important factor to change rill erosion, either morphology or erosion amount. Surface water content is high during spring thaw period because snowmelt and thawing of frozen soil water, which makes it prone to be washed. The obvious difference between snowmelt-induced rill erosion and the others is the existence of freeze-thaw action, presented as FT temperature, FT numbers, and thaw depth in our study. Among three factors above, thaw depth is the primary, as the unthawed frozen layer could promote rill erosion. Results also show that soil properties could also affect rill erosion in view of morphology and erosion amount. Accurate representation of freeze-thaw action has always been a problem in the lab, because in the field effect of the freeze-thaw action could be more complicate than our imagination. Meanwhile, snowmelt runoff has its unique characteristic compared to rain-induced runoff which most study focused. However, the results above are wished to provide certain reference to the future research, such as snowmelt erosion mechanism and appropriate control measures.

Acknowledgements

H. Fan has designed this research and completed some parts of this paper. Y. Liu has analyzed the rill-related data and completed some parts of this paper. X. Xu has designed the analysis method and completed the main parts of this paper. M. Wu has completed the experiment about the black soil. L. Zhou has completed the experiment about the albic soil.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes on contributors

Haoming Fan, Ph.D., is a professor at College of Water Conservancy, Shenyang Agricultural University, People’s Republic of China. He published articles on soil and water conservation and watershed management fields, focusing on freezing and thawing action, snowmelt erosion, gully erosion, and sediment yield budget in watersheds.

Yujia Liu is a postgraduate student at College of Water Conservancy, Shenyang Agricultural University, People’s Republic of China. Her major research is snowmelt-induced rill erosion.

Xiuquan Xu, Ph.D., is a lecturer at College of Water Conservancy, Shenyang Agricultural University, People’s Republic of China. His main research directions are snowmelt erosion characteristics at various spatial and temporal scales.

Min Wu, Ph.D., is a lecturer at College of Water Conservancy, Shenyang Agricultural University, People’s Republic of China. His main research interests are soil and water erosion and ephemeral gully erosion.

Lili Zhou, Ph.D., is an associate professor in College of Water Conservancy, Shenyang Agricultural University, People’s Republic of China. Her main research interest includes soil N and P cycles in cold environment.

Additional information

Funding

This research was supported by the National Natural Science Foundation of China [grant number 41371272].